BMI--just the same old crap with a new formula?

Factual Questions
1

This week I have seen two articles about Body Mass Index. It is not new. For those not familiar, it is calculated by the formula

(weight in pounds * 703)/(height in inches^2)

The result is a normalized index. If your index is over 25, you’re overweight.

This was supposedly a great improvement over the old height/weight charts, but it seems like it’s exactly the same, just using a formula so you don’t have to carry a chart around (but you have to remember that 703 that normalizes the index).

Anyway, here’s my point. There was a poster up at a gym I used to belong to. It showed two men who were both exactly the same height and weight. They would therefore also have the same BMI. But one was buff and lean and the other was fat. How can BMI mean anything at all?

2

As an afterthought I did a search and found two old threads. Here’s the better one. There may be nothing to add, but your’re welcome to.

3

For anyone who doesn’t work out, isn’t too old, and isn’t too short, the BMI correlates well with fatness. It is a mistake to apply it to everyone, but for the people who it works for, it works pretty well.

4

If you wonder about the 703 in there, it’s (inches/metre)[sup]2[/sup]/(kg/pound), which works out as closer to 704. (If you use SI units, there’s no factor at all, BMI is defined as (weight in kg) / (length in m)[sup]2[/sup].

Its great virtue is as a statistical tool. As you have noticed, it’s not always accurate for individuals, but if you look at a large population, it’s surprisingly effective, and you can say things like [I’m making this up] “people with a BMI > 27 are three times more likely of cardiovascular problems, compared to people with BMI < 18”